Cécile Mourer-Chauviré & Denis Geraads, the Struthionidae And

Home , Odontopteryx, Osteodontornis, Pelagornis, Pelagornithidae, Struthionidae

The Struthionidae and Pelagornithidae (Aves: Struthioniformes, Odontopterygiformes) from the late Pliocene of Ahl Al Oughlam, Morocco

Cécile Mourer-Chauviré 1 & Denis Geraads 2

1 : UMR 5125 PEPS CNRS ; Université Lyon 1, Campus de la Doua, Bât. Géode 69622 Villeurbanne Cedex, France. E- mail: [email protected] 2 : UPR 2147, Dynamique de l’Evolution Humaine, 44 rue de l’Amiral Mouchez, 75014 Paris, France. E-mail: denis. [email protected]

ABSTRACT - The Pliocene locality of Ahl al Oughlam is situated at the southeastern limit of the city of Casablanca, in Morocco, on an ancient seashore of the Atlantic Ocean. It has yielded a very rich vertebrate fauna (macro- and micromam- mals, birds, reptiles, amphibians and fishes) including both terrestrial and marine forms. On the basis of biostratigraphy, the fauna has been dated at about 2.5 Ma, which corresponds to the latest Pliocene. The avifauna is very diverse and includes birds belonging to twelve different orders. In this paper we describe only the Struthionidae and the Pelagornithidae. Ostriches are represented by a large-sized form, referred to the extinct species Struthio asiaticus Milne-Edwards, and its eggshells, of struthioid type, are comparable to those of the recent species Struthio camelus, but thicker. The Pelagornithidae, giant marine birds with bony pseudoteeth, are represented by an extinct species of the genus Pelagornis. As far as we know, this species was probably the latest representative of the order Odontopterygiformes.

Keywords - Aves, Ostriches, Pseudodontorns, Pelargornis mauretanicus n. sp., Pliocene, Morocco

Les Struthionidae et les Pelagornithidae (Aves, Struthioniformes et Odontopterygiformes) du Pliocène final d’Ahl al Oughlam, Maroc – Le gisement pliocène d’Ahl al Oughlam est situé à la limite sud-est de la ville de Casablanca, au Maroc, sur un ancien rivage de l’Océan Atlantique. Il a livré une très riche faune de Vertébrés (macro- et micromammifères, oiseaux, reptiles, amphibiens et poissons), appartenant surtout à des formes terrestres, mais comportant aussi des formes marines. Des arguments biostratigraphiques ont permis de dater cette faune de 2,5 Ma, ce qui correspond au Pliocène final. L’avifaune est très diversifiée et comporte des oiseaux appartenant à douze ordres différents. Dans cette étude nous décrivons seulement les Struthionidae et les Pelagornithidae. Les autruches sont représentées par une forme de grande taille, attribuée à l’espèce éteinte Struthio asiaticus Milne-Edwards, et leurs coquilles d’œufs, de type struthioïde, sont com- parables à celles de l’espèce actuelle Struthio camelus, mais sont plus épaisses. Les Pelagornithidae, oiseaux marins géants à pseudodents, sont représentés par une espèce éteinte du genre Pelagornis. Les restes trouvés à Ahl al Oughlam correspondent probablement aux derniers réprésentants connus de l’ordre des Odontopterygiformes.

Mots Clés - Oiseaux, Autruches, Oiseaux à pseudodents, Pelagornis mauretanicus n. sp., Pliocène, Maroc

INTRODUCTION a cliff on the shore.The fact that this site was formed along the oceanic seashore explains that both marine and terrestrial The fossiliferous locality of Ahl al Oughlam was forms were found among the birds. The mammals include discovered in 1985 by J.-P. Raynal and J.-P. Texier. It is situ- mainly terrestrial forms but the remains of a walrus and of ated at the southeast boundary of the city of Casablanca, Mo- two small cetaceans were also found. rocco, 6.5 km from the present-day Atlantic shore (Raynal et The biostratigraphical study of the mammals made al., 1990). This locality was then excavated by Denis Ger- it possible to date this locality to the Late Pliocene and to aads, within the framework of the “Programme Casablanca” attribute to it an age of about 2.5 Ma (Geraads, 1997). The of the “Institut National des Sciences de l’Archéologie et du mammals are very numerous and include more that 55 spe- Patrimoine” (INSAP), under the direction of Fatima-Zohra cies. The list of the identified taxa is given by Geraads, 1993, Sbihi-Alaoui and Jean-Paul Raynal. All the material comes 1995, 1996, 1997, 2004, 2006; Geraads et al., 1998; Geraads from a network of fissures and interconnected galleries, in a & Amani 1998; Alemseged & Geraads, 1998; Geraads & jumble of calcarenites blocks, at the foot of what was once Metz-Muller, 1999; and Eisenmann & Geraads, 2007.

ORYCTOS vol. 7, 2008 169 Figure 1 - Location map of the Ahl al Oughlam locality.

In addition to mammals and birds, fishes, amphib- AaO 2928; shaft of left tarsometatarsus, AaO 2216; shaft of ians, and reptiles have also been found. The amphibians and right tarsometatarsus, juv., AaO 2217; left tarsometatarsus, the reptiles other than chelonians have been studied by S. distal part, juv., AaO 2215; phalanx 1, AaO 836, and 2, AaO Bailon (2000). The material belongs to the collection of IN- 2219, of right posterior digit III; phalange 3 of posterior digit SAP, at Rabat. III, AaO 2973; phalanx 4 of posterior digit III, incomplete, Birds are very diverse and include Struthioni- AaO 877; 20 fragments of eggshells. formes, Procellariiformes, Odontopterygiformes, Pelecani- formes, Ciconiiformes, Anseriformes, Galliformes, Grui- Dimensions formes, Charadriiformes, Columbiformes, Psittaciformes, The material is very fragmentary and it is not al- Strigiformes, and Passeriformes. The present study concerns ways possible to take the standard measurements such as only the Struthionidae and Pelagornithidae. The anatomical width and depth of the proximal and distal parts, but when terminology generally follows Baumel & Witmer (1993), taking partial measurements and comparing them with the and Howard (1929) when necessary. The following museum same measurements taken on recent Struthio camelus, it is acronyms are used: BMNH, The Natural History Museum, found that they are on average 20 % larger than on a large London (formerly, British Museum (Natural History)); male individual (fig. 2). MNHN, Muséum National d’Histoire Naturelle, Paris; UCB, The measurements which could be taken are as fol- Université Claude Bernard, Lyon; USNM, collections of the lows (in mm): former United States National Museum, now in the National Proximal femur (AaO 873): maximum proximal Museum of Natural History, Washington, D.C. width: 134; minimum depth of collum femoris: 51 ; depth of caput femoris: 57.5. SYSTEMATIC PALEONTOLOGY Proximal tarsometatarsus (AaO 2928): proximal width: estimate superior to 92; width of the shaft at the level Order STRUTHIONIFORMES (Latham, 1790) of the distal end of the tuberosities for M. tibialis cranialis: Family STRUTHIONIDAE Vigors, 1825 ca. 47; depth of the shaft at the same level: ca. 51. Genus STRUTHIO Linnaeus, 1758 Distal tarsometatarsi AaO 2218 and AaO 2860: width of the shaft 10 cm proximad to the distal foramen 39.2; Struthio asiaticus Milne-Edwards, 1871 depth of the shaft at the same level: 27.8; AaO 2218: distal fig. 2-3 width at the level of the distal foramen: 64.0. The measurements of the phalanges are given in Material table 1. Dorsal vertebra and fragment of dorsal vertebra, AaO 835, 1444; 3 fragments of pelvis, AaO 703, 874, 875; Description fragment of rib, AaO 876; several pieces of a left femur, AaO Very numerous fossil forms of ostriches have been 873; left femur, distal part, AaO 700; shaft of left femur, juv., described from the Late Miocene, the Pliocene, and the AaO 3468; right femur, distal part, AaO 701; fragment of Pleistocene of Eurasia, both on osteological and eggshell right femur, distal part, AaO 702; left tibiotarsus, distal part, remains. According to Kurochkin & Lungu (1970), and AaO 1443; left tibiotarsus, distal part, juv., AaO 2214; right Mikhailov (1992), all the European and Asiatic Pliocene tarsometatarsus, distal part, and left tarsometatarsus, distal forms, described from osteological remains, are very similar part, from the same individual, trochleae not preserved, AaO and have been synonymized with the species Struthio asi- 2860, 2218; left tarsometatarsus, proximal part, incomplete, aticus Milne-Edwards, described from the Pliocene of the

ORYCTOS vol. 7, 2008 170 Figure 2 - Struthio asiaticus, Ahl al Oughlam. a-b: left femur, proximal part, AaO 873; a) caudal view, b) cranial view. Same scale for a and b. c-d: phalanx 1 of right posterior digit III, distal part, AaO 836; c) dorsal view, b) plantar view. e-f: phalanx 2 of right posteri- or digit III, AaO 2219; e) dorsal view, f) plantar view. Same scale for c, d, e, and f.

Siwaliks (Milne-Edwards, 1867-71). However, according to All the species described from osteological remains Mikhailov (1992: p. 370) “the eggshell remains… are dis- from the Late Miocene, the Pliocene, and the Early Pleis- tinct and the features of each “eggshell morphotype species” tocene of Europe and Asia, with the exception of S. orlovi, evolved independently from the Miocene to the Pleistocene”. from the Late Miocene of Moldavia (Kurochkin & Lungu, The “eggshell species” need to have their own names, at 1970), are massive, heavy ostriches, characterized by dimen- least parataxonomic ones. For that reason the eggshells are sions and robustness larger than those of the recent ostrich, S. referred to three different species: S. asiaticus, which has camelus. This also applies to the ostriches which have been struthioid type eggshells, S. chersonensis, which has mixed described after the publication of Kurochkin and Lungu’s aepyornithoid-struthioid type eggshells, and S. transcauca- (1970) work, for example S. dmanisensis, from the Early sicus, which has aepyornithoid type eggshells (Mikhailov, Pleistocene of Georgia (Burchak-Abramovich & Vekua, 1988, 1992). 1990). Although S. asiaticus is not taller than the recent os-

ORYCTOS vol. 7, 2008 171 Figure 3 - Struthio asiaticus, Ahl al Oughlam. Scanning electron microscope pictures of eggshell fragments showing the complexes of pore openings: a-b: eggshell n° 7; c: eggshell n° 8; d-f: eggshell n° 20. For the eggshell fragment n° 20, the pore openings are particularly visible because of deposits of Mn salts inside them. (Scale bars = 0.1 mm) trich, it is also a massive form, as shown by the proportions of the phalanx 1 of posterior digit III (table 2). Comparisons with the African forms It is not always possible to compare all these forms In Africa Struthio oldawayi was described by Lowe with each other since they have been described on the ba- (1933) from fragments of pelvis coming from Olduvai, Bed sis of different elements, but for a certain number of spe- I, Tanzania, dated from 2 - 2.1 to 1.7 Ma (Leakey, 1967; Hay cies, the dimensions of phalanx 1 of posterior digit III are 1976). Lowe indicates that the fossil form only differs from known (Sauer, 1979). For S. asiaticus, the dimensions of the recent subspecies of S. camelus by its larger size but does the proximal part of this phalanx are known (Davies, 1880; not give its dimensions. Leakey (1967: pl. 96 and 97 ) illus- Lydekker, 1884-1886) and for the Ahl al Oughlam ostrich trates some bones of a very large ostrich, also coming from the dimensions of the distal part are known. For all these fos- Olduvai, Bed I. These bones are at least 33 % larger than sil ostriches, the ratio between proximal width and minimum those of a large recent male ostrich. Later on, Arambourg width of shaft is comprised between 1.57 and 1.68, whereas (1979) described the species Struthio barbarus from the Vil- in the recent S. camelus it varies from 1.68 to 1.92. Likewise lafranchian locality of Aïn Boucherit, in Algeria, also char- the ratio distal width/minimum width of the shaft is com- acterized by its dimensions that are larger than those of the prised between ca 1.37 and 1.58 in the fossil ostriches, while recent forms. This locality is dated from the Plio-Pleistocene it varies from 1.54 to 1.86 in the recent ones (table 2). For boundary and thus it is slightly younger than Ahl al Oughlam S. asiaticus, as well as for the Ahl al Oughlam ostrich, the (Geraads, 1993). The dimensions of S. barbarus are on aver- measurements of this phalanx indicate that they were mas- age 20 % larger than those of a large recent male ostrich. sive ostriches. However, among the forms which have been synonymized

ORYCTOS vol. 7, 2008 172 with S. asiaticus, some are up to 50 % larger than the recent lus, which are generally from 1.6 to 2.1 mm thick, with an ones (Kurochkin & Lungu, 1970). According to the differ- exception for the recently extinct subspecies S. c. syriacus, ent types of fossil eggshells, there were different forms of in which eggshell thickness was from 1.3 to 2.1 mm (Sauer, ostriches in Africa during the Pliocene and the Early Pleis- 1966). tocene (aepyornithoid-type eggshells at least until the Mio- The eggshells attributed to Struthio asiaticus and Pliocene, and several different struthioid-type eggshells). coming from Nullas, on the Ken river, Banda district, in In- Due to the lack of information concerning the type of egg- dia, also show the same type of pore openings as the recent shells of S. oldawayi and S. barbarus, it is not possible to subspecies S. camelus molybdophanes, and their thickness know if they were really different from each other. varies from 2.2 to 2.4 mm (Bidwell, 1910; Sauer, 1968).

Comparisons with fossil eggshells Comparison with the other fossil eggshells de- Introduction scribed from Africa The difference between the aepyornithoid- and the 1° Eggshells from the continental Mio-Pliocene of struthioid-type eggshells lies in the arrangement and the Morocco surface openings of the pore canal systems. “The Aepyornis Sauer and Sauer (1978) have described struthious eggs have simply isolated canals, each of which opens to eggshells found in Mio-Pliocene continental deposits of the surface in its own elongate and sometimes forked pore southern Morocco. These eggshells show pore openings groove” (Sauer, 1966: p. 10). Also in the Aepyornis eggs, which are aepyornithoid, struthioid, or mixed A-S. It is “the deeply grooved pore openings [are] oriented parallel, sometimes possible to find two different types on the same and interspersed with short comma and sting (or “dagger eggshell, aepyornithoid pore openings in the equatorial re- point”) pores” (Sauer, 1966: p. 11). “In the [recent] Struthio- gion, and struthioid pore openings in the polar region. There nidae the eggs have multi-branched pore canals with multi- is some correlation between the thickness of the shell and the ple openings in roundish pits” (Sauer, 1966: p. 10). type of pore openings. The eggshells with struthioid pores As mentioned above, the eggshells of the fossil os- are generally from 1.55 to 2 mm thick and the eggshells with triches belonging to the genus Struthio show pore openings aepyornithoid pores are generally from 1.95 to 2.7 mm thick, which can be aepyornithoid, mixed, or struthioid. In the re- but there are many exceptions in both groups. Sauer and cent forms of Struthio camelus the eggshells are of struthioid Sauer think that there is an evolution between the aepyorni- type, but the different subspecies show different types of pore thoid-type eggshells and the struthioid-type eggshells, with openings. Five types of pore openings have been defined for all the intermediates between them. However the struthioid- the five subspecies S. c. camelus, S. c. spatzi, S. c. austra- type eggshells of these deposits have a type of pore openings lis, S. c. massaicus and S. c. molybdophanes (Schönwetter, which resembles those of the recent subspecies S. c. camelus 1927; Sauer, 1972). However Sauer (1972) indicates that and S. c. australis, and are therefore different from the Ahl al the association between the subspecies and the type of pore Oughlam eggshells. openings is not always as constant as indicated by Schönwet- ter (1927). 2° Eggshells of the genus Psammornis The genus Psammornis, based on eggshells, in- Description of the Ahl al Oughlam eggshells (fig. cludes two species: P. rothschildi, from eggshells found in 3) Algeria, and P. lybicus, from one fragment found in Lybia. On the eggshells from Ahl al Oughlam, when they In addition, some fragments found in Arabia have been re- have not been altered by erosion or by diagenesis, one can ferred to Psammornis sp. This material was revised by Sauer see multiple pore openings grouped together in shallow pits. (1969). Each pit includes about 50 openings. The diameter of the pits These eggshells are thick and the Algerian and varies from 0.5 to 1 mm and their number varies from 10 to Arabian forms correspond to large-sized eggs. According 20 per cm2, depending on the fragments. This structure is to Sauer the original thickness of the eggshell is 4.0 mm close to that of the recent subspecies S. c. molybdophanes, in P. rothschildi, 3.1 mm in Psammornis sp., and 2.80 mm which lives in Somalia, in the south-east and south of Ethio- in P. lybicus. The size of the egg is estimated at about 250 pia, the south-east of Sudan, and in Kenya (Mayr & Cottrell, x 190 mm in P. rothschildi, 210 x 172 mm in Psammornis 1979), and differs from that of the subspecies which live in sp., and 143 x 127 in P. lybicus. Their age is not accurately geographically closer areas, such as S. c. camelus (South known but according to Sauer (1969), the size of the eggs of Sahara and Mauritania, and from Senegal to Sudan and and the thickness of the shells suggest that they are Pliocene. Eritrea), or S. c. spatzi which used to live in the Rio de Oro The arrangement of the pore openings is typically struthioid region, but which is now merged with S. c. camelus (Brown and resembles more particularly that of the subspecies S. c. et al., 1982). camelus. The eggshells of the Ahl al Oughlam ostrich are At Ahl al Oughlam the thickness of the eggshell distinct from the different forms of Psammornis; they are not fragments varies from 2.3 to 2.7 mm, with an average value so thick and the disposition of the pore openings corresponds of 2.54. These eggshells are thicker that those of S. came- rather to that of S. c. molybdophanes.

ORYCTOS vol. 7, 2008 173 3° Eggshells from Namibia Conclusion Very numerous eggshell localities have been dis- It seems important to remind here that the species covered in Namibia and have made it possible to establish Struthio coppensi, from the Early Miocene of Namibia, a biostratigraphy based on these stratigraphically collected which is a true didactylous ostrich, was associated with eggshells (Pickford et al., 1995; Senut & Pickford, 1995; aepyornithoid eggshells (Mourer-Chauviré et al., 1996). The Senut et al., 1998; Pickford & Senut, 2000; Senut 2000). presence of aepyornithoid eggshells does not mean that the Several eggshell localities have also yielded fossil mammals birds which laid them were Aepyornithidae. In absence of “which permit the ages of the strata to be determined within skeletal remains it is not possible to write about an aepyor- reasonable limits” (Senut & Pickford, 1995, p. 33). This bi- nithoid eggshell : “This record confirms that elephant birds ostratigraphy demonstrates the succession of three genera, survived in mainland Africa until at least the Late Miocene” Namornis, from the Middle Miocene (approx. - 16 to - 15 (Harrison & Msuya, 2005, p. 313). Ma), then Diamantornis, with four species from the Middle Two different types of ratite eggshells are found si- and Late Miocene (approx. - 15 to - 8 Ma), then the genus multaneously in the Neogene of Africa and Southern Eura- Struthio with three species, S. karingarabensis, from the sia: Late Miocene (approx. - 8 to - 5 Ma), S. daberasensis, from –eggshells with relatively simple aepyornithoid, the Pliocene (approx. - 5 to - 2 Ma), and lastly S. camelus, or struthioid of solitary needle-point type, or intermediate from the Early Pleistocene to the Recent (Pickford & Senut, aepyornithoid-struthioid openings, such as those described 2000; Senut, 2000). by Sauer & Sauer (1978) in the Mio-Pliocene continental de- In the genus Diamantornis the pore openings are posits of Morocco. In these forms, pore complexes including clustered in very large numbers to form megapores, circular a large number of openings are not found. In East Africa, in surface area, and funnel-shaped in section, the diameter of aepyornithoid eggshells are found until the Late Miocene, at which varies from 2.7 to 10 mm according to the species; the about 6.5 to 6.7 Ma (Harrison & Msuya, 2005). shell thickness varies from 2.1 to 4.2 mm. –struthious eggshells with large pore complexes In the two extinct species of the genus Struthio, grouping together a large number of openings, as described the pore openings are clustered to form pore complexes but in the four species of the genus Diamantornis, and in the they are distinctly smaller than in the genus Diamantornis. three extinct species Struthio karingarabensis, S. kakesien- In Struthio karingarabensis the diameter of the pore com- sis, and S. daberasensis. These eggshells have been first de- plexes varies from 2.4 to 2.7 mm and the thickness of the scribed in Namibia but they are also present in South Africa shell varies from 2.95 to 3.2 mm. In S. daberasensis the (Pickford & Senut, 2000), East Africa (Malawi, Tanzania, diameter of the pore complexes varies from 0.5 to 2.2 mm Kenya), and in Arabia (Harrison & Msuya, 2005; Stidham, and their number varies from 4 to 5 per cm2. The shell thick- 2004 a). These large pore complexes are still found in one ness is situated between 1.7 and 2.6 mm, with a mode of 2.3 of the subspecies of the recent ostrich, S. camelus molyb- mm. The Ahl al Oughlam eggshells are close to those of S. dophanes. daberasensis by their thickness but differ from them by the At Ahl al Oughlam the diameter and the density of generally smaller size of the pore complexes (0.5 to 1mm) the pore complexes are similar to that of this recent subspe- and by their larger density (10 to 20 per cm2). cies, but the thickness of the shell is larger, and closer to that of the extinct species S. daberasensis. S. daberasensis 4° Eggshells from East Africa is found in deposits dating from approximately 2 to 5 Ma in Harrison & Msuya (2005) have described struthio- Namibia, and in deposits dating from ca. 3.0 to 3.5 in Ma- nid eggshells from the localities of Laetoli, in Tanzania, lawi. The eggshells found at Ahl al Oughlam are thus con- Lothagam and Kanapoi, in Kenya, the ages of which are sistent with the chronological position of this locality. comprised between 7.4 and ca. 2.6-2.7 Ma. They have found Struthionid skeletal remains associated with egg- again the chronological succession of two species of Dia- shells have rarely been found in the same fossiliferous local- mantornis, then Struthio cf. karingarabensis, S. kakesiensis, ity. The case of Ahl al Oughlam is particularly interesting a new species, morphologically intermediate between S. because it shows the association of skeletal remains of a mas- karingarabensis and S. daberasensis from Namibia, then S. sive ostrich, Struthio asiaticus, with thick eggshells, thicker camelus which appears in East Africa at approximately 3.6 than those of the recent S. camelus, and the pore openings - 3.8 Ma. Aepyornithoid eggshells have also been found at of which are similar to those of the recent subspecies S. c. Lothagam, in a layer dated from ca. 6.5 to 6.7 Ma, thus in molybdophanes. the Late Miocene, while in Namibia aepyornithoid eggshells are only found in Early Miocene and early Middle Miocene Order ODONTOPTERYGIFORMES Howard, layers (Pickford & Senut, 2000; Senut, 2000). 1957 At last, in Malawi, struthioid eggshells referred to Family PELAGORNITHIDAE Fürbringer, 1888 S. daberasensis have been reported in the Mid-Pliocene (ca. 3.0 - 3.5 Ma) of the Chiwondo Beds ( Stidham, 2004 a). The Pelagornithidae are members of a group of extinct birds which can be designated as pseudodontorns.

ORYCTOS vol. 7, 2008 174 They were giant marine birds which presented bony tooth- cording to B. Cahuzac (pers. comm.) the Pelagornis from shaped excrescences on the jaws. These pseudoteeth are Condom could also come from the “Molasse coquillère ma- unique among birds. They had a very large wingspan, up to rine” of the Middle Miocene, and its age would be the same 5.5 to 6 m, and their skeleton was highly modified. They had as that of the holotype. This material is conserved in the Mu- a superficial similarity with the Procellariiformes but they seum of Bordeaux collection and casts are kept in the collec- were more closely related to the Pelecaniformes. However, tions of BMNH, London, and UCB, Lyon. recent works on the skull of very early pseudodontorns show that they form a monophyletic group and that they are not Pelagornis mauretanicus n. sp. included in the Pelecaniformes (Bourdon, 2004). It is thus fig. 4-10 preferable to place them in their own order, the Odontop- terygiformes Howard, 1957. A large number of forms have Holotype. been described (Harrison & Walker, 1976; Olson, 1985), but, Right humerus, distal part, AaO 705 (fig. 4, a-b) according to Olson and Rasmussen (2001: p. 259): “So far, all of the considerable material of these birds found in late Horizon and Locality. Oligocene and Neogene deposits around the North Atlantic Late Pliocene, age about 2.5 Ma. Ahl al Oughlam, appears to be referable to a single genus - Pelagornis - al- south-east boundary of the city of Casablanca, Morocco. though this includes a variety of species “. The pseudodontorns are known since the Late Pale- Paratypes. ocene of Central Asia (Averianov et al., 1991) and the Early Mandible, fragments of the right and left ramus, Eocene of England (Harrison & Walker, 1976) and Morocco AaO 884; fragments of rostrum, AaO 881, 882, 896; frag- (Gheerbrant et al., 2003; Bourdon, 2006). As early as this, ments of rostrum or mandible with pseudoteeth, AaO 880, they were probably present on all the oceans because they 883; isolated pseudoteeth, AaO 711; right quadrate, AaO are also known in the Late Eocene or the Early Oligocene 707; right pterygoids, AaO 708 et 878; vertebrae, AaO 879, of Antarctica (Tonni, 1980; Tonni & Tambussi, 1985). They 886, 888, 889, 890, 893, 898, 2620; fragment of sacrum, were widespread during the Miocene but they are more AaO 887; fragments of proximal left humerus, AaO 895, rarely reported from later deposits. During the Pliocene they proximal right humerus, AaO 710, distal right humerus, AaO are present in North America, as shown by material from the 706; right ulna, distal part, right radius, distal part, and right Yorktown Formation, at Lee Creek, North Carolina (Olson os carpi radiale, associated, AaO 709; left ulna, distal part, & Rasmussen, 2001), dated from 4.5 to 5 Ma. They have also and left radius, distal part, associated, AaO 894; right radius, been reported from New Zealand (Mc Kee, 1985), where proximal part, AaO 885; fragment of right radius, distal part, new discoveries show them to be present up to layers dated AaO 899; right os carpi ulnare, AaO 712; fragment of left from 3.6 to 3.1 Ma (Mc Kee, pers. comm.). In Japan Ono carpometacarpus, proximal part, AaO 891; fragment of wing (1980) described a femur from the Upper Pliocene Dainichi phalanx, AaO 821; right femur, almost complete, AaO 897. Fm., Kakegawa Group of Shizuoka Pref., which was later at- All the pseudodontorn material from Ahl al Ough- tributed to a pseudodontorn (Matsuoka et al., 1998). In Peru, lam has been referred to a single species, P. mauretanicus, in the Pisco Formation, pseudodontorns are present in the lo- because all these elements are compatible in size. This mate- cality of Sacaco, dated from about 3 Ma (Chavez & Stucchi, rial consists mainly in isolated bones, and it has rarely been 2002). The Moroccan form, dated ca. 2.5 Ma, is therefore the found several representatives of the same bone. For the dis- most recent, accurately dated, form known so far. tal part of the radius, however, there are three elements, the dimensions of which are similar (table 9). When it is pos- Genus PELAGORNIS Lartet, 1857 sible to compare with other pseudodontorns, e.g. those from the Lee Creek Mine (Olson & Rasmussen, 2001), the Ahl al This genus only includes one species, Pelagornis Oughlam form is consistently slightly larger than Pelagornis miocaenus Lartet, 1857, and this species is only known by sp. 2, and clearly larger than Pelagornis sp. 1 (tables 5 and humeri and a shaft of ulna. The holotype is an almost com- 10). plete left humerus coming from the “Molasse coquillère marine de l’Armagnac”, department of Gers. Its age is Middle Diagnosis. Miocene, Serravallian (Cahuzac, pers. comm.), equivalent Species comparable in size to Pelagornis miocae- of the mammalian biozones MN 6-7-8, age between 16.5 nus. Processus supracondylaris dorsalis projecting both dor- and 11 Ma (BiochroM’97, 1997). It has been illustrated by sally and cranially, and clearly separated by a groove from Milne-Edwards (1867-71: pl. 45) and it is kept in the Paris the condylus dorsalis. This processus extends in proximal MNHN, uncatalogued. The other elements attributed to this direction by a wing which obliquely merges with the cranio- species come from the localities of Léognan, near Bordeaux, dorsal angle of the shaft. Two depressions are present in the and Condom, in Gers. The locality of Léognan (Cestats) is fossa m. brachialis, one proximal to the condylus dorsalis, Early Miocene, Burdigalian or Orleanian, MN 3 zone, age and the other one, deeper, proximal to the condylus ventra- between 20 and 18 Ma (Duranthon & Cahuzac, 1997). Ac- lis. The epicondylus ventralis is strongly developed cranio-

ORYCTOS vol. 7, 2008 175 Figure 4 - Pelagornis mauretanicus n. sp., Ahl al Oughlam. a-b: right humerus, distal part, AaO 705, holotype; a) cranial view, b) caudal view; c-d: left humerus, proximal part, AaO 895, paratype; c) cranial view, d) caudal view; e-f: right pterygoid, AaO 708, paratype; e) dorsal view, f) lateral view; g-h: right pterygoid, AaO 878, paratype; g) dorsal view, h) lateral view; i-j: right quadrate, AaO 707, paratype; i) medial view, j) lateral view.

ORYCTOS vol. 7, 2008 176 Figure 5 - Diagram showing the way of measuring the distal part of the humerus of Pelagornis mauretanicus n. sp. The hatched area corresponds to an incompletely preserved part.

caudally. constant characteristic because it is hardly visible on another right humerus from Léognan (original in Bordeaux, cast in Dimensions: see tables 4-10. Lyon n° O. 758), as well as on a humerus attributed to the genus Pelagornis and coming from New Zealand (Scarlett, Etymology. 1972; Harrison & Walker, 1976: pl. 10, fig. C). On the latter From Mauretania, name of the ancient Roman prov- two humeri, the distal outline of the caput humeri only forms ince, which encompassed a part of present-day Morocco. a slight expansion. This expansion is situated between two depressions, the first one corresponds to the incisura capitis, Curation of the material. and the second one is a fossa situated distally compared to Institut National des Sciences de l’Archéologie et the tuberculum dorsale, but it does not form “a distal ridge du Patrimoine (INSAP), at Rabat, Morocco. with marked internal curvature”. The two proximal parts of humerus of Ahl al Oughlam, although they do not present a Description and comparisons recurved distal ridge, can nevertheless be attributed to the genus Pelagornis. Harrison and Walker have given an emended diagnosis of the humerus of Pelagornis. They indicate that at the Comparison of the humeri with those of Pelagornis proximal part, on the caudal face: “a deep hollow extends miocaenus (fig. 4, a-d) up towards external tuberosity and another hollow undercuts The Ahl al Oughlam remains are slightly larger than the proximal head, leaving a distal ridge, with marked inter- the holotype of P. miocaenus but these differences are not nal curvature, on the anconal surface” (Harrison & Walker, significant because, as already mentioned by Milne-Edwards 1976: p. 29). This can also be expressed by saying that the (1874), there are large size differences between the different caput humeri is extended distally by a ridge, recurved in humeri attributed to this species (table 3). The humerus from ventral direction and situated between two depressions. This Condom is almost 25 % larger than the holotype humerus, or ridge is clearly visible on the holotype left humerus from the Léognan humerus (cast n° A 167, BMNH). Armagnac, illustrated by Milne-Edwards (1867-71: pl. 45) The distal parts of humerus from Ahl al Oughlam and on the right humerus from Léognan illustrated by Har- differ from those of P. miocaenus by the following charac- rison and Walker (1976: pl. 10, fig. A et B) but it is not a ters: in P. miocaenus the processus supracondylaris dorsalis

ORYCTOS vol. 7, 2008 177 Figure 6 - Diagram of quadrates of Pelagornithidae. a-b: Pelagornis mauretanicus n. sp., Ahl al Oughlam, right quadrate, AaO 707; a) medial view, b) cranial view; c-d: Pelagornis sp. 1 or Pelagornis sp. 2, Yorktown Fm., Lee Creek Mine, right quadrate, USNM 446495; c) medial view, d) craniolateral view; e: indetermined Pelagornithidae, Calvert Fm., right quadrate, USNM 237189, cranial view. The hatched areas correspond to incompletely preserved parts.

is weakly projecting, both on the dorsal and cranial faces, in In distal view, due to the great extension of the epicondy- the Ahl al Oughlam form it is more projecting both dorsally lus ventralis, the Ahl al Oughlam humerus is craniocaudally and cranially, and it is clearly separated from the condylus larger than the P. miocaenus humerus (fig. 5 and table 4). dorsalis by a groove. Proximally it is extended by a wing Considering these morphological differences and which obliquely merges with the craniodorsal angle of the the great interval of time (17.5 to 8.5 Ma) which separates shaft. In P. miocaenus the surface of the shaft situated proxi- them, we attribute the Ahl al Oughlam form to a new spe- mally to the condyles is flattened, while in the Ahl al Ough- cies. lam form it shows two fossae, one proximal to the dorsal condyle, the other, deeper, proximal to the ventral condyle. Description of the other elements In P. miocaenus the ventral condyle is slightly flattened in the proximodistal direction, while in the Ahl al Oughlam form it A. Quadrate AaO 707 (fig. 4, i-j, fig. 6 and table 5) is more rounded. On the caudal face, the fossa olecrani is On this specimen the tip of the processus orbitalis absent but the epicondylus ventralis, although incomplete in is broken. The processus oticus ends in two condyles, the P. miocaenus as well as in P. mauretanicus, was probably capitulum oticum and the capitulum squamosum, separated more developed and caudally projecting, and was probably by a distinct incisura intercapitularis. The capitulum squa- extending further proximally in the Ahl al Oughlam form. mosum is rounded, the capitulum oticum is formed by two

ORYCTOS vol. 7, 2008 178 Figure 7 - Diagram of pterygoids of Pelagornis ssp. a-b: Pelagornis sp. 1 or Pelagornis sp. 2, Yorktown Fm., Lee Creek Mine, right pterygoid, USNM 425110; a) dorsomedial view, b) cranial view; c-d: Pelagornis mauretanicus n. sp., Ahl al Oughlam, right pterygoid, AaO 708; c) dorsomedial view, d) cranial view.

flat surfaces, one medially oriented and the other one crani- which is also broken. It differs from that from Ahl al Ough- ally oriented, joined along an oblique edge. Distally to the lam because the capitulum oticum is not formed of two flat capitulum oticum, on the cranial face, the dorsal side of the sides but rounded, and because it does not show a triangular processus orbitalis forms a triangular surface, moderately shallow surface on the proximal side of the processus orbit- deep, and bordered by two thin ridges. Distally to this trian- alis, distally to the capitulum oticum. It was also compared gular surface, on the medial side of the processus orbitalis, with the quadrate of Pelagornis sp., from the Early Pliocene there are two large pneumatic foramina. The cotyla quadra- of Lee Creek (USNM 446 495, Olson & Rasmussen, 2001: tojugalis is socket-shaped and surrounded by a thick rim. It pl. 11, f), the processus orbitalis of which is not preserved. is situated at about the distal third of the bone. The processus The Ahl al Oughlam quadrate is extremely similar to the lat- mandibularis is formed by two rounded protuberances, the ter one, which also shows a capitulum oticum formed by two condylus medialis and the condylus lateralis, separated by flat sides. The only difference is that in medial view the Lee an oblique groove. The condylus pterygoideus is short and Creek quadrate shows a pneumatic foramen situated proxi- looks like a synostosed articulation. mally to the articular mandibular surface, at the level of the This quadrate was compared with that of a Pel- groove which separates the condylus medialis from the con- agornithidae from the Calvert Formation, dated from the dylus lateralis. At the location of this pneumatic foramen Middle Miocene (USNM 237 189), the processus orbitalis of there is only a shallow depression on the Ahl al Oughlam

ORYCTOS vol. 7, 2008 179 ORYCTOS vol. 7, 2008 180 quadrate. Oughlam, as well as the Lee Creek one, have the same gen- A pseudodontorn quadrate, designated as Osteo- eral shape as that of Odontopteryx. In the latter, the facies dontornis sp., was described from the Middle Miocene of articularis basipterygoidea is well developed and elongate, Japan (Ono, 1989). It also looks very similar in shape and and there is a point cranially to the facies articularis palatina size to that from Ahl al Oughlam but it only shows a small but this point is much more projecting in Odontopteryx. pneumatic foramen, on the medial face, at the level of the Aslanova and Burchak-Abramovich (1999: p. 428) middle part of the processus orbitalis. mention the presence of the pterygoid of Caspiodontornis Among the recent Pelecaniformes the quadrate of and indicate that it strongly differs from that of Odontop- Pelagornis looks more similar to that of Sulidae and Pha- teryx and looks more similar to that of the recent genus Pe- ethontidae. The characteristics shared with these two fami- lecanus. However, in Pelecanus, the pterygoid is narrow and lies are a socket-shaped articulation for the quadrato-jugal, a elongate and there is no basipterygoid facet. In addition, the processus oticus showing two distinct condyles, a pneumatic shape of the facies articularis palatina is very different from foramen situated on the medial face, dorsally compared to that of Pelagornis and Odontopteryx, it is concave and dor- the condylus pterygoideus and at the same level as the ventral soventrally elongated. It seems therefore that the pterygoid edge of the processus orbitalis. It differs however because in of Caspiodontornis was different from that of the other pseu- the Sulidae and Phaethontidae the quadrato-jugal socket is dodontorns. situated at the end of a peduncle which extends laterally and In all the recent Pelecaniformes, the perygoid has ventrally, while this is not the case in Pelagornis. In addition, the shape of a narrow, elongate stick, triangular in section in the shape of the mandibular articulation is very different in the Pelecanidae, circular or subcircular in section in the Pha- Pelagornis from that of the Sulidae and Phaethontidae. The ethontidae and Fregatidae, laterally flattened in section in the characteristics in common between Pelagornis, the Sulidae Phalacrocoracidae, Anhingidae and Sulidae. The pterygoid and the Phaethontidae are also found in the Diomedeidae of Odontopteryx, and even more that of Pelagornis, differs and, in the shape of the mandibular articulation, the quadrate from that of the recent Pelecaniformes by its short and cra- of Pelagornis also resembles that of Diomedea. nially wide shape. In the recent Diomedeidae the pterygoid also has the shape of a narrow, elongate stick and the facies B. Pterygoids AaO 708 and 878 (fig. 4, e-h, fig. 7 articularis basipterygoidea is semicircular and it is situated and table 5) almost at the cranial end of the bone. By these two char- The two pterygoids differ slightly from each other: acteristics the pterygoid of Pelagornis differs from that of AaO 878 is larger and the facies articularis quadratica shows Diomedeidae. a rounded cotyla and a pneumatic foramen while on AaO 708 this articular facet is a little synostosed with bony ex- C. Mandible and pseudoteeth (fig. 8-9) crescences around it and does not show that pneumatic fo- The mandible AaO 884 includes the two mandibu- ramen. The facies articularis basipterygoidea is quadrangular lar rami resting against each other and covered by sand and in AaO 708, elongate and more elliptical in AaO 878. The concretioned sediments. It is not possible to free them be- ventral wing is slightly projecting in AaO 708, and more pro- cause the bones are completely crushed. It is possible to see, jecting in AaO 878. The facies articularis palatina shows a on both rami, the position of the intraramal joint (Zusi & pneumatic foramen on both specimens, it is more circular in Warheit, 1992). A part of the posterior ramus of the right AaO 708 and more elongate in AaO 878. mandible is preserved over 9 cm posteriorly to the intraramal The pterygoid AaO 878 is very similar to a pterygoid joint, and cranially to the joint the right and left intermediate of Pelagornis sp. from the Pliocene of Lee Creek (USNM rami are preserved over 18 cm, but the pseudoteeth are not 425 110, Olson & Rasmussen 2001: pl. 11, d). On this speci- preserved on the anteriormost parts. The part corresponding men there is a projecting point at its cranial end, cranially to the mandibular symphysis is absent. On the lateral face to the facies articularis palatina. This point was probably of the left ramus the mandibular lateral longitudinal sulcus present in AaO 878 but has been broken. It does not exist in (Harrison & Walker, 1976), or external groove (Stidham, AaO 708 where there is at the same place a slightly raised 2004 b) is clearly visible. crest but no point. The pseudoteeth are always arranged according to In pseudodontorns, the pterygoid was described in the same pattern. The largest ones, of order 1, are separated Odontopteryx toliapica, from the Early Eocene of England by about 40 mm (table 6). In the middle of this space there are (Harrison & Walker, 1976). Both pterygoids from Ahl al smaller ones, of order 2. In the middle of the space between

Figure 8 (opposite) - Pelagornis mauretanicus n. sp., Ahl al Oughlam. a-e: mandible AaO 884, paratype; a) right lateral view, b) left lateral view, c) detail of the pseudoteeth of the right ramus, caudally to the break of the mandible, d) closer detail of pseudoteeth of the right ramus. From the left to the right: pt. order 1, needle not preserved, pt. order 3, needle, pt. order 2, needle, pt. order 3, needle, e) closer detail of pseudoteeth of the left ramus, cranial part. From the left to the right: pt. order 1, needle, pt. order 3, needle, pt. order 2, needle, pt. order 3, needle. f: fragment of rostrum AaO 882, paratype. From the left to the right: pt. order 1, needle, pt. order 3, needle, pt. order 2, needle.

ORYCTOS vol. 7, 2008 181 Figure 9 - Diagram showing the position of order 1, 2 and 3 pseudoteeth and needles, on the right and left rami of the AaO 884 mandible. The needles in dotted lines have been broken but their trace is visible on the mandible. Their is a small discrepancy, of 1 or 2 mm, between the position of the pseudoteeth on the right and left rami. This can be due to the imprecision of the measurement, or to a slight postdeposi- tional displacement. The pseudoteeth are slightly closer together on the caudal part than on the cranial part.

Figure 10 (opposite) - Pelagornis mauretanicus n. sp., Ahl al Oughlam. a-b: left ulna, distal part, and left radius, distal part, AaO 894, paratype; a) ventral view, b) cranial view; c: right ulna, distal part, right radius, distal part, and right radial carpal, AaO 709, paratypes. Ulna and radius, ventral view, radial carpal, distal view; d: right ulnar carpal, AaO 712, paratype, medial view; e-f: left carpometacarpus, proximal part, AaO 891, paratype; e) dorsal view, f) cranioventral view. Same scale for a, b, c, d, e, and f. g-h: right femur, AaO 897, paratype; g) cranial view, h) caudal view of the distal part; i: jumble of bones showing a large mammal tooth and the proximal part of a left radius of P. mauretanicus n. sp., AaO 885, paratype, caudal view. On the left, there are a left radius and a left ulna, distal parts, crushed. Same scale for g, h, and i. the order 1 and order 2 teeth are smaller ones, of order 3. ramus, but on its lateral side. The measurements of the pseu- Finally, in the middle of the space between the teeth of order doteeth on the mandible AaO 884 are given in table 7. 1 or 2 and the teeth of order 3 there are extremely thin pseu- The order 1 and 2 pseudoteeth are not regularly el- doteeth, which are truly needles of bone. These pseudoteeth liptical in section. They show a ridge situated either on the are not always perfectly preserved, but even when they are laterocaudal angle, or on the mediocaudal angle. The order broken it is possible to see their location on the dorsal face 3 pseudoteeth and the needles are craniocaudally flattened, of the mandible. For the needles the trace is a thin line. The their width is larger than their depth. The whole set of pseu- needles are not situated quite on the top of the mandibular doteeth is oriented perpendicularly compared to the longitu-

ORYCTOS vol. 7, 2008 182 ORYCTOS vol. 7, 2008 183 dinal axis of the mandible, and they are not inclined crani- in proximal direction. These two condyles are practically in- ally. distinguishable from the surface of the shaft. On the contrary The last two visible pseudoteeth, of order 3, are sit- the tuberculum carpale is strongly separated from the shaft uated either cranially, or just caudally, to the intraramal joint. and proximodistally elongated. There is a depression on the Caudally to these two last ones, there are no more traces of dorsal face of the tuberculum carpale. Dimensions (mm): pseudoteeth. diagonal width from the caudalmost side of condylus dorsa- The pattern of the pseudoteeth corresponds to what lis to tuberculum carpale, 24.4; depth of condylus dorsalis: has been described by Harrison and Walker (1976) for the 21.7. largest part of the mandible of the genus Osteodontornis. It differs from the pattern of a mandible illustrated by Zusi and F. Radii (fig. 10, a-b, i, and table 9) Warheit (1992) for an indetermined Pelagornithidae (ChM The proximal part of the radius (AaO 885) shows PV 4768) from the Charleston Museum. However almost the an oval, dorsoventrally elongated proximal articular surface same pattern can be found on a partial rostrum of Pelagorni- and, on each side, a strong tuberculum bicipitale radii. The thidae (probably Osteodontornis orri) from the Miocene of distal part of the radius is ventrally incurved and forms a kind California illustrated by Olson (1985: p. 195, fig. 9) and on of niche in which the distal part of the ulna comes to lodge. a dentary of Osteodontornis sp. from the late Early Miocene This disposition is reminiscent of what can be observed for of Japan (Matsuoka et al., 1998). New material of Osteodon- example in frigatebirds, but the hollow of the radius is more tornis orri has been found in the Middle Miocene of Shark- pronounced in Pelagornis. As in Fregatidae, the distal part of tooth Hill, in California (Stidham, 2004 b). Its age is between the radius is inflated cranially, while in other Pelecaniformes 15.2 and 16.0 Ma, which corresponds to the beginning of (Pelecanidae, Phaethontidae, some Phalacrocoracidae, Suli- the Middle Miocene. The arrangement of the teeth on the dae) the distal part of the radius is widened, like a paddle, mandible is slightly different from what can be observed on but not ventrally incurved and cranially inflated. In the ge- the Ahl al Oughlam mandible, but this may be due to the fact nus Anhinga, however, it is slightly cranially inflated, but not that this fragment corresponds to a more anterior part than ventrally incurved. that preserved on the Moroccan mandible. The Ahl al Oughlam pattern is also different from G. Ulnar carpal AaO 712 (fig. 10, d) that which was described for Caspiodontornis kobystani- This bone is relatively narrow and the two crura are cus, from the Middle Oligocene of Azerbaijan (Aslanova & not very divergent. The crus longum is not much longer than Burchak-Abramovich, 1999), where the succession of large, the crus breve and there is a large pneumatic foramen in the medium, and small “teeth”, separated by a variable number incisura metacarpalis. Among the recent Pelecaniformes it of “toothlets”, is not as regular as in the Moroccan form. In is most similar to the Pelecanidae, but in the latter the bone addition, in the Azerbaijan form, certain pseudoteeth of the is much more pneumatized. It strongly differs from that of mandible are slightly cranially inclined. the Diomedeidae in which the crus longum is much longer than the crus breve. Dimensions (mm): total length or length D. Fragments of rostrum (fig. 8, f) of crus longum: 21.2; length of crus breve: 16.0; maximum The fragment of rostrum AaO 896 is completely width at the top of the two crura: 21.1; width at the base: crushed and it is not possible to see many details except for 18.2. the edge of both tomia with incomplete pseudoteeth. On the lateral side of the left tomium one can see the rostral lateral H. Radial carpal AaO 709 (fig. 10, c) longitudinal sulcus, or rostral groove, and on the crushed This bone has an overall trapezoidal shape, with a palate one can see a deep tooth pit. Fragments AaO 881 flattened and almost semicircular distal surface. It is higher and 882 show pseudoteeth of order 1, 2, and 3, incomplete, cranially than caudally. In its general shape it is reminiscent and needles. They are considered as fragments of rostrum of the Pelecanidae, Phalacrocoracidae, Phaethontidae and because they show a rostral groove which is situated at a Sulidae, and differs from the Fregatidae and Diomedeidae. small distance from the tomial edge. On these fragments the Dimensions (in mm): maximum width of the distal articular dimensions of the pseudoteeth are smaller than on mandible surface: 20.8; maximum depth of the distal articular surface: AaO 884 (table 8). These fragments perhaps correspond to 14.0; proximodistal length on the cranial side: 15.0; proxi- a more anterior part of the jaws than the part preserved on modistal length on the caudal side: 10.5. mandible AaO 884. I. Proximal carpometacarpus AaO 891 (fig. 10, e-f) E. Distal ulnae (fig. 10, a-c) The alular metacarpal is incomplete but it is possi- The two distal ulnae AaO 709 and AaO 894 seem to ble to see that it was situated very far distally from the proxi- come from the same individual. The distal articular surface is mal extremity, and that it was very slightly projecting in rather flat while it is usually rounded in birds, both condyles, proximal direction. On the caudal face, distally to the troch- dorsalis and ventralis, are very slightly projecting on the lea carpalis, the os metacarpale minus is not separated from ventral face, and the condylus dorsalis is weakly elongated the os metacarpale majus, which indicates that the two met-

ORYCTOS vol. 7, 2008 184 acarpal bones were fused proximally over a long distance. tionary history in Africa, as evidenced by the occurrence of This shape is reminiscent of that in the genus Pelecanus, but the oldest representative of the genus Struthio, S. coppensi, in the latter the fovea carpalis cranialis is more developed, in the Early Miocene of Namibia (Mourer-Chauviré et al., the median groove between the two ridges, dorsal and ven- 1996), and by the presence of several eggshell species of the tral, of the trochlea carpalis, is deeper, and these two ridges genus Struthio in the Late Miocene and Pliocene of Southern are much more projecting, while in Pelagornis the median and Eastern Africa. groove is hardly indicated. Dimensions (mm): dorsoventral Among the Odontopterygiformes, the genus Pel- width of the trochlea carpalis at the proximal end: 17.5. agornis was previously known only by humeri and by a small number of other remains from the locality of Lee Creek J. Femur (fig. 10, g-h, and table 10) (Olson & Rasmussen, 2001). The material found at Ahl al The femur AaO 897 was crushed in the locality, its Oughlam made it possible to associate it with the humeri and shaft was entirely plastered and it is partly covered by con- to describe a large part of the mandible and pseudoteeth, and cretioned sediments. It is not very different from a femur of other elements of the postcranial skeleton. It is possible that recent Pelecanus but its collum femoris is not so contracted, some of the other genera of Odontopterygiformes will turn and its trochanter femoris is not so projecting caudally, so out to be synonymous with the genus Pelagornis, which was that the proximal articular surface has a more rectangular the first genus to be described in this order. outline. The distal part of this femur is widened and flattened ACKNOWLEDGMENTS as in the recent Pelecanidae, and it is especially widened on the lateral side. On the cranial face the two condyles are very The authors are grateful to J. Hassar-Benslimane, slightly projecting and the sulcus patellaris is wide and shal- head of the INSAP, for having entrusted them with the study low. On the caudal face the Ahl al Oughlam femur differs of the birds from Ahl al Oughlam. We thank Bruno Cahuzac, from that of a recent Pelecanus because the sulcus fibula- from the University of Bordeaux I, for the information con- ris (fibular groove of Howard, 1929) is hardly indicated and cerning the age of the French Pelagornis localities. Cécile thus the lateral and fibular condyles are practically united. Mourer-Chauviré is very grateful to Helen James and Storrs The part situated proximally to the condyles (fossa poplitea) Olson for their hospitality in Washington, D. C., and for the seems to have been rather flat but it is partly masked by plas- possibility of studying the fossil material of the Calvert Fm. ter. and of the Lee Creek Mine. She also thanks her colleagues Among the pseudodontorns the femur is known Ursula Göhlich, Gerald Mayr, and Martin Pickford for help- in Palaeochenoides miocaenus (Shufeldt, 1916; Wetmore, ful discussions. We both thank Noël Podevigne, from the 1917) the age of which can be considered as Late Oligocene Université Claude Bernard - Lyon 1, for taking the photo- (Olson 1985). The Ahl al Oughlam femur differs from it be- graphs. cause in P. miocaenus the condylus lateralis is very clearly separated from the condylus fibularis. REFERENCES The femur of Pelagornis mauretanicus is very similar to that of Pelagornis sp., small form, from the Pliocene of Alemseged, Z. & Geraads, D. 1998. Theropithecus atlanticus the Yorktown Fm. of the Lee Creek Mine (Olson & Rasmus- (Thomas, 1884) (Primates : Cercopithecidae) from the sen, 2001: pl. 11, p-t et q-u). In this form the collum femoris late Pliocene of Ahl al Oughlam, Casablanca, Morocco. is also weakly contracted but the trochanter femoris seems to Journal of Human Evolution, 34: 609-621. be more caudally projecting. The distal part is also very flat- Arambourg, C. 1979. Vertébrés villafranchiens d’Afrique du tened on the cranial face, with a wide, shallow sulcus patel- Nord (Artiodactyles, Carnivores, Primates, Reptiles, laris. The caudal face of the distal part is eroded but it seems Oiseaux). Fondation Singer-Polignac ed., Paris: 1-141. that the sulcus fibularis, between the lateral and fibular con- Aslanova, S. M. & Burchak-Abramovich, N. I. 1999. A dyles, was also very weakly indicated. detailed description of Caspiodontornis kobystanicus from the Oligocene of the Caspian seashore. Acta CONCLUSIONS zoologica cracoviensia, 42 (3): 423-433. Averianov, A. O., Panteleyev, A. V., Potapova, O. R. & Nessov, Although this study only concerns two families L. A. 1991. Bony-toothed birds (Aves: Pelecaniformes: from the Ahl al Oughlam avifauna, it provides interesting Odontopterygia) of the late Paleocene and Eocene of the systematic and paleobiogeographical information. Western margin of ancient Asia; pp. 3-12. In Potapov, The ostrich is referred to the species Struthio asiati- R. L. (ed.) Ecology and Fauna of the Eurasiatic Birds. cus, which was widespread in a large part of Eurasia during USSR Academy of Sciences, Proceedings of the the Pliocene. This does not mean, however, that the ostriches Zoological Institute, 239 (in Russian). originated in Eurasia and reached Africa at a recent period, Bailon, S. 2000. Amphibiens et reptiles du Pliocène terminal as has sometimes been proposed, e. g. by Mikhailov (1992), d’Ahl al Oughlam (Casablanca, Maroc). Geodiversitas, or Rasmussen et al. (2001). Ostriches have a long evolu- 22 (4): 539-558.

ORYCTOS vol. 7, 2008 185 Baumel, J. J. & Witmer, L. M. 1993. Osteologia; pp. Geraads, D. 1996. Le Sivatherium (Giraffidae, Mammalia) du 45-132. In Baumel, J. J. (ed.) Handbook of Avian Pliocène final d’Ahl al Oughlam (Casablanca, Maroc) Anatomy. Nomina anatomica avium. 2nd edition. Nuttall et l’évolution du genre en Afrique. Paläontologische Ornithological Club, 23. Zeitschrift, 70 (3-4): 623-629. Bidwell, E. 1910. Remarks on some fragments of egg-shell Geraads, D. 1997. Carnivores du Pliocène terminal de Ahl of a fossil Ostrich from India. Ibis, ser. 9, 4: 759-761. al Oughlam (Casablanca, Maroc). Geobios, 30 (1): BiochroM’ 97. 1997. Synthèse et tableaux de corrélations; 127-164. pp. 769-805. In Aguilar, J. P., Legendre, S. & Michaux, Geraads, D. 2004. New skulls of Kolpochoerus J. (eds.) Actes du Congrès BiochroM’97. Mémoires et phacochoeroides (Suidae: Mammalia) from the late Travaux de l’ Ecole pratique des hautes études, Institut Pliocene of Ahl al Oughlam, Morocco. Palaeontologia de Montpellier, 21. africana, 40: 69-83. Bourdon, E. 2004. A new Paleocene-Eocene avifauna from Geraads, D. 2006. The late Pliocene locality of Ahl al Morocco: the early diversification of the pseuo-toothed Oughlam, Morocco: vertebrate fauna and interpretation. birds (Aves, Odontopterygiformes). Sixth international Transactions of the Royal Society of South Africa, 61 meeting of the Society for Avian Paleontology and (2): 97-101. Evolution, Quillan, France, 28th September-3rd October, Geraads, D. & Amani, F. 1998. Bovidae (Mammalia) du 2004. Abstracts: 9. Pliocène final d’Ahl al Oughlam, Casablanca, Maroc. Bourdon, E. 2006. L’avifaune du Paléogène des phosphates Paläontologische Zeitschrift, 72 (1/2): 191-205. du Maroc et du Togo. Diversité, systématique et apports Geraads, D., Amani, F., Raynal, J.-P. & Sbihi-Alaoui, F.-Z. à la connaissance de la diversification des oiseaux 1998. La faune de Mammifères du Pliocène terminal modernes (Néornithes). Thèse de Doctorat du Muséum d’Ahl al Oughlam, Casablanca, Maroc. Comptes national d’Histoire naturelle, Paris, 187 p., 28 pl., 5 Rendus de l’Académie des Sciences, Paris, Sciences de annexes. la terre et des planètes, 326: 671-676. Brown, L. H., Urban, E. K. & Newman, K. 1982. The Birds Geraads, D. & Metz-Muller, F. 1999. Proboscidea of Africa. Vol. I. Academic Press ed., London and New (Mammalia) from the late Pliocene of Ahl al Oughlam York: I-XIII + 1-521. (Casablanca, Morocco). Neues Jahrbuch Geologie Burchak-Abramovich, N. I. & Vekua, A. K. 1990. The fossil Paläontologie Monatshefte, 1999 (1): 52-64. ostrich Struthio dmanisensis sp. n., from the Lower Gheerbrant, E., Sudre, J., Cappetta, H., Mourer-Chauviré, C., Pleistocene of Georgia. Acta zoologica cracoviensia, Bourdon, E., Iarochène, M., Amaghzaz, M. & Bouya, 33 (7): 121-132. B. 2003. Les localités à mammifères des carrières Chavez, M. & Stucchi, M. 2002. El registro de Pelagornithidae de Grand Daoui, bazssin des Ouled Abdoun, Maroc, (Aves: Pelecaniformes) en el Pacifico sudeste. 1° Yprésien : premier état des lieux. Bulletin de la Société Congreso Latinoamericano de Paleontologia de géologique de France, 174 (3):279-293. Vertebrados, Santiago de Chile, Octubre 2002. Harrison, C. J. O. & Walker, C. A. 1976. A review of the bony- Davies, W. 1880. On some fossil bird-remains from the toothed birds (Odontopterygiformes): with descriptions Siwalik Hills in the British Museum. The Geological of some new species. Tertiary research special paper, Magazine, New Series, Decade II, 7: 18-27. London, 2: 1-62. Duranthon, F. & Cahuzac, B. 1997. Eléments de corrélation Harrison, T. & Msuya, C. P. 2005. Fossil struthionid eggshells entre échelles marines et continentales : les données from Laetoli, Tanzania: Taxonomic and biostratigraphic du Bassin d’Aquitaine au Miocène; pp. 591-608. In significance. Journal of African Earth Sciences, 41: J.-P. Aguilar, J.-P.; Legendre, S. & Michaux, J. (eds.) 303-315. Actes du Congrès BiochroM’97. Mémoires et. Travaux Hay, R. L. 1976. Geology of the Olduvai Gorge: A study de l’Ecole pratique des hautes études, Institut de of sedimentation in a semiarid basin. University of Montpellier, 21. California Press, Berkeley: 1-203. Eisenmann, V. & Geraads, D. 2007. Hipparion pomeli Howard, H. 1929. The avifauna of the Emeryville shellmound. sp. nov. from the late Pliocene of Ahl al Oughlam, University of California Publications in Zoology, 32: Morocco, and a revision of the relationships of Pliocene 301-394. and Pleistocene hipparions. Palaeontologia africana, Howard, H. 1957. A gigantic “toothed” marine bird from 42: 51-98. the Miocene of California. Santa Barbara Museum of Geraads, D. 1993. Kolpochoerus phacochoeroides (Thomas, Natural History, Department of Geology, 1: 1-23. 1884) (Suidae, Mammalia) du Pliocène supérieur de Kurochkin, E. N. & Lungu, A. N. 1970. A new Ostrich from Ahl al Oughlam (Casablanca, Maroc). Geobios, 26 (6): the Middle Sarmatian of Moldavia. Paleontological 731-743. Journal (translation of Paleontologicheski Zhurnal), Geraads, D. 1995. Rongeurs et Insectivores du Miocène final 1970, 1: 103-111. de Ahl al Oughlam, Casablanca, Maroc. Geobios, 28 Lartet, E. 1857. Note sur un humérus fossile d’Oiseau (1): 99-115 attribué à un très grand palmipède de la section

ORYCTOS vol. 7, 2008 186 des Longipennes. Comptes Rendus des séances de Birds from the Lee Creek Mine, North Carolina; pp. l’Académie des Sciences de Paris, XLIV: 736-741. 233-365. In Ray, C. E. & Bohaska, D. J. (eds.) Geology Leakey, L. S. B. 1967. Olduvai Gorge. 1951-1961. Vol. and Paleontology of the Lee Creek Mine, North Carolina, 1. A preliminary report on the Geology and Fauna. III. Smithsonian Contribution to Paleobiology, 90. Cambridge University Press: 1-118. Ono, K. 1980. Pliocene tubinare bird from Kakegawa, Lowe, P. R. 1933. On Some Struthious Remains: - 1. Shizuoka Prefecture, Japan. Memoirs of the National Description of some Pelvic Remains of a large Fossil Science Museum, 13: 11-15. Ostrich, Struthio oldawayi, sp. n., from the Lower Ono, K. 1989. A Bony-Toothed Bird from the Middle Miocene, Pleistocene of Oldaway (Tanganyika Territory); 2. Chichibu Basin, Japan. Bulletin of the National Science Egg-shell Fragments referable to Psammornis and Museum, Series C, Geology & Paleontology, Tokyo, 15 other Struthiones collected by Mr. St. John Philby in (1): 33-38. Southern Arabia. Ibis, 1933: 652-658. Pickford, M. & Senut, B. 2000. Geology and Paleobiology of Lydekker, R. 1884-1886. Indian Tertiary and Post-Tertiary the central and southern Namib Desert, Southwestern Vertebrata. Vol. III, part 4. Siwalik Birds. Palaeontologia Africa, vol. 1: Geology and History of study. Geological Indica, Calcutta, ser. X, 3: 135-147. Survey of Namibia, Memoir 18: 1-155. Matsuoka, H., Sakakura, F. & Ohe, F. 1998. A Miocene Pickford, M., Senut, B. & Dauphin, Y. 1995. Biostratigraphy pseudodontorn (Pelecaniformes: Pelagornithidae) from of the Tsondab sandstone (Namibia) based on gigantic the Ichishi Group of Misato, Mie Prefecture, Central avian eggshells. Geobios, 28 (1): 85-98. Japan. Palaeontological Research, 2 (4): 246-252. Rasmussen, D. T., Simons, E. L. & Hertel, F. 2001. Hindlimb Mayr, E. & Cottrell, G. W. (Eds.) 1979. Check-list of of a giant terrestrial bird from the Upper Eocene, Birds of the World. Vol. I, second edition. Museum of Fayum, Egypt. Palaeontology, 44 (2): 325-337. Comparative Zoology, Cambridge Mass.: I-XVII + Raynal, J.-P., Texier, J.-P., Geraads, D. & Sbihi-Alaoui, 1-547. F.-Z. 1990. Un nouveau gisement paléontologique McKee, J. W. A. 1985. A pseudodontorn (Pelecaniformes: plio-pléistocène en Afrique du Nord : Ahl Al Oughlam Pelagornithidae) from the middle Pliocene of Hawera, (ancienne carrière Deprez) à Casablanca (Maroc). Taranaki, New Zealand. New Zealand Journal of Comptes Rendus de l’Académie des Sciences, Paris, Zoology, 1985, 12: 181-184. 310, Série II: 315-320. Mikhailov, K. E. 1988. The comparison of East European Sauer, E. G. F. 1966. Fossil Eggshell Fragments of a Giant and Asian ostriches pliocene eggshells. Fossil Reptiles Struthious Bird (Struthio oshanai, sp. n.) from Etosha and Birds of Mongolia. Transactions of the Joint Soviet- Pan, South West Africa. Cimbebasia, 14: 2-51. Mongolian Palaeontological Expedition, 34. Moscow, Sauer, E. G. F. 1968. Calculations of Struthious egg sizes Nauka: 65-72 (in Russian, English summary). from measurements of shell fragments and their Mikhailov, K. E. 1992. The microstructure of Avian and correlations with phylogenetic aspects. Cimbebasia, Dinosaurian eggshell: Phylogenetic implications; pp. ser. A, 1 (2): 27-55. 361-373. In Campbell, K. E. Jr. (ed.) Papers in Avian Sauer, E. G. F. 1969. Taxonomic Evidence and Evolutionary Paleontology honoring Pierce Brodkorb. Natural Interpretation of Psammornis. Bonner zoologische History Museum of Los Angeles County, Science Series, Beiträge, 20 (1/3): 290-310. 36. Sauer, E. G. F. 1972. Ratite Eggshells and Phylogenetic Milne-Edwards, A. 1867-71. Recherches anatomiques et Questions. Bonner zoololgische Beiträge, 23 (1): 3-45. paléontologiques pour servir à l’histoire des oiseaux Sauer, E. G. F. 1979. A Miocene Ostrich from Anatolia. Ibis, fossiles de la France. Victor Masson et Fils édit., 121: 494-501. Paris: t. I, 1-474; atlas, pl. 1-96 ; t. II, 1-627; atlas, pl. Sauer, E. G. F. & Sauer, E. M. 1978. Ratite Eggshell Fragments 97-200. from Mio-Pliocene continental Sediments in the District Milne-Edwards, A. 1874. Observations sur les oiseaux of Ouarzazate, Morocco. Palaeontographica, Abt. A, fossiles des faluns de Saucats et de la molasse de 161: 1-54. Léognan. Bibliothèque de l’Ecole des Hautes Etudes Scarlett, R. J. 1972. Bone of a presumed odontopterygian (Section des Sciences Naturelles), 11 (3): 3-12. bird from the Miocene of New Zealand. New Zealand Mourer-Chauviré, C., Senut, B., Pickford, M. & Mein, P. 1996. Journal of Geology and Geophysics, 15 (2): 269-274. Le plus ancien représentant du genre Struthio (Aves, Schönwetter, M. 1927. Die Eier von Struthio camelus spatzi Struthionidae), Struthio coppensi n. sp., du Miocène Stresemann. Ornithologische Monatsberichte, XXXV inférieur de Namibie. Comptes Rendus de l’Académie (1): 13-17. des Sciences de Paris, 322, sér. II a: 325-332. Senut, B. 2000. Fossil ratite eggshells: A useful tool for Olson, S. L. 1985. The fossil record of birds; pp. 79-252. In Cainozoic biostratigraphy in Namibia. Communications Farner, D. S., King, J. R. & Parkes, K. C. (eds.) Avian of the geological Survey of Namibia, 12: 367-373. Biology, 8, Academic Press, New York. Senut, B. Dauphin, Y. & Pickford, M. 1998. Nouveaux Olson, S. L. & Rasmussen, P. C. 2001. Miocene and Pliocene restes aviens du Néogène de la Sperrgebiet (Namibie):

ORYCTOS vol. 7, 2008 187 complément à la biostratigraphie avienne des éolianites Odontopterygia) de gran tamaño del Terciario temprano du désert de Namib. Comptes Rendus de l’Académie de Antartida. Ameghiniana, XVII (3): 273-276. des Sciences, Paris, Sciences de la terre et des planètes, Tonni, E. P. & Tambussi, C. P. 1985. Nuevos restos de 327: 639-644. Odontopterygia (Aves: Pelecaniformes) del Terciario Senut, B. & Pickford, M. 1995. Fossil eggs and Cenozoic temprano de Antartida. Ameghiniana, 21 (2-4): continental Biostratigraphy of Namibia. Palaeontologia 121-124. africana, 32: 33-37. Wetmore, A. 1917. The relationships of the fossil bird Shufeldt, R. W. 1916. New extinct bird from South Carolina. Palaeochenoides miocaenus. The Journal of Geology, Geological Magazine, new. ser., 3: 343-347. 25: 555-557. Stidham, T. A. 2004 a. Extinct ostrich eggshell (Aves: Zusi, R. L. & Warheit, K. E. 1992. On the evolution of the Struthionidae) from the Pliocene Chiwondo Beds, intraramal mandibular joints in Pseudodontorns (Aves: Malawi: implications for the potential biostratigraphic Odontopterygia); pp. 351-360. In Campbell, K. E. Jr. correlation of African Neogene deposits. Journal of (ed.) Papers in Avian Paleontology honoring Pierce Human Evolution, 46: 489-496. Brodkorb. Natural History Museum of Los Angeles Stidham, T. A. 2004 b. New skull material of Osteodontornis County, Science Series, 36. orri (Aves: Pelagornithidae) from the Miocene of California. PaleoBios, 24 (2): 7-12. Tonni, E. P. 1980. Un pseudodontornitido (Pelecaniformes, Autorisation de publier n° 4/04 du 24 Janvier 2005

ORYCTOS vol. 7, 2008 188 Ph. 1 D III Ph. 2 D III Ph. 3 D III Ph. 4 D III AaO 836 AaO 2219 AaO 2973 AaO 877

Total length - 74.0 35.3 - Length as preserved 87.4 - - 67.0 Proximal width - 47.7 35.8 31.7 Proximal depth - 38.5 26.4 27.2 Width in the middle 32.3 31.0 37.2 - Depth in the middle 25.0 24.0 18.0 - Width of the distal art. surf. 47.4 ca. 44.5 39.0 - Depth of the distal art. surf. ca. 39.7 28.3 19.5 -

Table 1 - Struthio asiaticus, from Ahl al Oughlam, dimensions (mm) of the phalanges of the posterior digit III.

Phalanx 1 of post. digit III Proximal Minimum Ratio Distal Ratio width (1) width of (1):(2) width (3) (3):(2) shaft (2)

S. cf. brachydactylus 41.2 25.3 1.63 - - S. cf. karatheodoris 48.O 28.5 1.68 44.0 1.54 S. brachydactylus 40.0 25.0 1.60 39.5 1.58 S. sp. Maragha - 29.0 - 45.0 1.55 S. asiaticus 41.9 25.0 1.67 - - S. pannonicus 55.0 est. 35.0 ca. 1.57 ca. 48.0 ca. 1.37 AaO 836 - 31.2 - 47.4 1.52

Mean of fossil Sruthio 1.63 (n=5) 1.51 (n=5)

Recent S. camelus 43.8 23.8 1.84 41.5 1.74 41.3 21.5 1.92 40.0 1.86 42.5 22.3 1.90 40.4 1.81 38.3 21.6 1.77 35.8 1.66 37.8 21.2 1.78 36.0 1.70 40.9 23.3 1.76 37.5 1.61 40.6 24.1 1.68 37.2 1.54 32.9 19.2 1.71 30.6 1.59 33.5 19.0 1.76 30.7 1.62 43.6 22.8 1.91 39.4 1.73

Mean of recent S. camelus 1.80 (n=10) 1.69 (n=10)

Table 2 - Comparisons of the dimensions (mm) and ratios of the phalanx 1 of the posterior digit III in fossil and recent species of Struthio.

ORYCTOS vol. 7, 2008 189 Table 1

Table 2 Pelagornis miocaenus

Humerus Holotype Léognan Léognan Condom Armagnac BMNH cast Bordeaux Bordeaux Paris MNHN n° A 167 Museum Museum (a) (b)

Total length sup. to 580 591 sup. to 625 est. 710

Proximal width 61 61.5 59.3 - as preserved as preserved as preserved

Width at the level of the bicipital surface 40 - - -

Largest diameter of shaft in the middle 22 - 24.6 28.0

Smallest diameter of shaft in the middle 19 - 19.8 25.0

Distal width distally to proc. suprac. dorsalis 50 - - -

Maximum width at distal end - 53.7 - -

Distal depth 39 37.6 * - -

Length from tuberc. dors. to the dist. end of deltoid crest 168 174 est. 195 est. 210

Table 3 - Comparisons of the dimensions (mm) of the humeri of Pelagornis miocaenus from the French Miocene localities (a) after the text and illustrations of Milne-Edwards (1867-71). (b) after Harrison & Walker (1976). *: “anterior to posterior width at internal condyle” (Harrison & Walker, 1976, p. 33).

ORYCTOS vol. 7, 2008 190 Table 3 Pelagornis mauretanicus Pelagornis miocaenus n. sp. Holotype Léognan Léognan Armagnac BMNH Bordeaux Humerus, proximal part AaO 895 AaO 710 Paris Cast n° A Museum MNHN 167 (a)

Depth of caput humeri 26.5 26.2 22.0 21.2 23.3 as preserved

Humerus, distal part AaO 705 AaO 706 Holotype Paratype

Distal width from the proc. supracond. dors. to the ventral 52 51.2 43.0 - - side of cond. ventr. (ﬁg. 5, A) as preserved as preserved

Distal width of the two condyles measured on the dist. art. surf. 46 - 40.0 - - (ﬁg. 5, B) as preserved

Distal width from the most dorsal part of cond. dors. to the protuberance 55 - 49.0 - - on the ventral face (ﬁg. 5, C) as preserved

Distal depth from the cranial face of cond. ventr. to the most caudal part of 51.5 - 39.0 - - epicond. ventr. (ﬁg. 5, D) as preserved

Distal depth at the level of cond. dors. (ﬁg. 5, E) 39.4 - 31.5 32.5 -

Proximodistal length of cond. ventr. 18.6 18.6 17.5 - - (ﬁg. 5, F) as preserved

Proximodistal length of cond. ventr. 32.0 30.5 28.0 - - (ﬁg. 5, G) as preserved

Table 4 - Comparisons of the dimensions (mm) of the humeri of Pelagornis mauretanicus n. sp., from Ahl al Oughlam, with those of Pelagornis miocaenus. (a) after Harrison & Walker (1976).

ORYCTOS vol. 7, 2008 191 Table 4 P. mauretanicus Pelagornis Indetermined Osteodontornis n. sp. sp. 1 or sp. 2 Pelagornithidae sp. Chichibu Quadrate AoO 707 Lee Creek Calvert Fm. Basin (a) (b)

Maximum height 48.4 45.6 44.1 44.6

Maximum depth, measured on the medial face, from cond. pter. to the caudalmost 31.7 - - 33.0 edge of cot. quadrat.

Maximum depth at the top of proc. otic. 18.0 - - 16.8

Maximum depth at the level of the surf. artic. mand. ca. 26 - - 27.2

Pelagornis mauretanicus Pelagornis n. sp. sp. 1 or sp. 2 Pterygoid AaO 708 AaO 878 Lee Creek (a)

Total length 42.9 sup. to 48 40.1

Greatest cranial width 14.6 16.9 17.2

Table 5 - Comparisons of the dimensions (mm) of the quadrate and pterygoids of Pelagornis mauretanicus n. sp., from Ahl al Oughlam, with those of other Pelagornithidae (a) after Olson & Rasmussen (2001). (b) after the text and illustrations of Ono (1989).

ORYCTOS vol. 7, 2008 Table 5 192 Pelagornis mauretanicus n. sp. mandible AaO 884

Right ramus Left ramus

Axis of pseudotooth order 2 0 Axis of needle 6 mm Axis of pseudotooth order 3 12 mm Trace of needle 17 mm Axis of pseudtooth order 1 24 mm Axis of pseudtooth order 1 0 Trace of needle 30.5 mm Trace of the needle 7 mm Axis of pseudotooth order 3 35 mm Axis of pseudotooth order 3 12 mm Axis of needle 39 mm Axis of needle 16 mm Axis of pseudotooth order 2 45 mm Axis of pseudotooth order 2 22 mm Axis of needle 50 mm Axis of needle 27 mm Axis of pseudotooth order 3 54 mm Axis of pseudotooth order 3 31 mm Axis of needle 58 mm Axis of needle 36 mm Axis of pseudotooth order 1 66 mm Axis of pseudotooth order 1 42 mm Break of the right ramus Break of the right ramus Axis of needle 72 mm missing part Axis of pseudotooth order 3 76 mm missing part Axis of needle 81 mm Axis of needle 59 mm Axis of pseudotooth order 2 86 mm Axis of pseudotooth order 2 64 mm Axis of needle 90 mm Trace of needle 68.5 mm Axis of pseudotooth order 3 94 mm Axis of pseudotooth order 3 72 mm Axis of needle 98 mm Trace of needle 74.5 mm Axis of pseudotooth order 1 103 mm Axis of pseudotooth order 1 80 mm Axis of needle 109 mm Axis of needle 87 mm Axis of pseudotooth order 3 114 mm Axis of pseudotooth order 3 92 mm Axis of needle 119 mm Trace of needle 95.5mm Axis of pseudotooth order 2 123 mm Axis of pseudotooth order 2 100 mm Axis of needle 129 mm Axis of needle 107 mm Intraramal joint Axis of pseudotooth order 3 112 mm Axis of pseudotooth order 3 136 mm Intraramal joint

Table 6 - Pelagornis mauretanicus n. sp., from Ahl al Oughlam, distance from the median axis of the pseudoteeth to that of the first preserved pseudotooth on the two rami of the mandible AaO 884.

Pelagornis mauretanicus n. sp. Height Craniocaudal length at the base Mandible AaO 884

Pseudoteeth of order 1 20 9 Pseudoteeth of order 2 11 5 Pseudoteeth of order 3 7 2.5 Needles 3 1

Table 7 - Pelagornis mauretanicus n. sp., from Ahl al Oughlam, dimensions (mm) of the pseudoteeth on the mandible AaO 884.

ORYCTOS vol. 7, 2008 193 Table 6

Table 7 Pelagornis mauretanicus n. sp. Table 8 - Pelagornis mauretanicus n. sp., from Ahl al Fragments of rostrum AaO 881 AaO 882 Oughlam, dimensions (mm) Height Craniocaudal Height Craniocaudal of the pseudoteeth on the length at the length at the fragments of rostrum. base base

Pseudoteeth of order 1 est. 16 8 - 7.4 Pseudoteeth of order 2 - - - 5 Pseudoteeth of order 3 - 3 - 2.7 Needles - 1 - 0.9

Pelagornis mauretanicus n. sp. Table 9 - Pelagornis mau- Radius AaO 885 AaO 894 AaO 899 AaO 709 retanicus n. sp., from Ahl al Oughlam, dimensions (mm) Proximal width 25.2 - - - of the fragments of radius. Proximal depth 24.6 - - - Distal width - 28.7 26.5 26.4 as preserved as preserved Distal depth - 21.5 - -

Pelagornis Pelagornis Pelagornis Palaeochenoides mauretanicus sp. 1 sp. 2 miocaenus n. sp. Lee Creek Lee Creek (b) Femur AaO 897 (a) (a)

Total length as preserved 133 - - -

Proximal width 32.4 29.0-30.0 - -

Proximal depth ca. 18 - - -

Depth of caput femori 16.0 - - -

Distal width 34.5 29.6-30.6- sup. to 33 40 30.9-32.7

Depth at the level of 18.2 - - ca. 25 condylus medialis

Depth at the level of 24.5 - - 25 condylus lateralis

Table 10 - Comparisons of the dimensions (mm) of the femur of Pelagornis mauretanicus n. sp., from Ahl al Oughlam, with those of other Pelagornithidae. (a) after Olson & Rasmussen (2001); (b) after Shufeldt (1916).

ORYCTOS vol. 7, 2008 194 Table 8

Table 9

Table 10